SAND Lab, University of Chicago

Super proud advisor moment today, as MIT Technology Review announced that Shawn Sixiong Shan, PhD student co-advised by me and Heather Zheng at Univ. of Chicago, is MIT TR Innovator of the Year, 2024, for his projects protecting human creativity.

Watching the announcement live on LinkedIn, was quite a surreal moment. And now with this added to his list of awards, Shawn is finally turning his attention to his faculty job applications for the coming cycle!

Unveiling the 2024 Innovator of the Year | LinkedIn Every year, MIT Technology Review recognizes 35 Innovators Under 35. These young entrepreneurs, researchers, and humanitarians are inventing materials and building systems to help tackle the world’s most pressing problems in biotechnology, computing, and climate science. On Monday, September 9, we...

I realized I posted this to a closed group, but should have also posted to my main timeline...

So weird to post things from Instagram here, but this painting by Eva is special. https://www.instagram.com/p/C0Uks1jrPLl/

The painting is called Belladonna "lumen inter tenebras," and it's the first painting processed by Nightshade.

Belladonna, "lumen inter tenebras". 🧪☠🖤 My new oilpainting for the group show Intertwined at @wowxwow_art is now available! This painting is extremely special to me. I will explain why! This year, I have been intertwined (pun intended) in fighting for artist rights against predatory AI companies. At the start of 2024, a new technology designed to protect artists will be released called Nightshade, created by the University of Chicago. Here is a short description of the tool by MIT Technology Review : "The tool, called Nightshade, is intended as a way to fight back against AI companies that use artists' work to train their models without the creator's permission. Using it to "poison" this training data could damage future iterations of image-generating AI models, such as DALL-E, Midjourney, and Stable Diffusion, by rendering some of their outputs useless—dogs become cats, cars become cows, and so forth." Over the past year, I have befriended Ben Y. Zhao, the lead scientist who created this technology (and also @theglazeproject) together with his team. I was so moved by his and his team's continual efforts in protecting artists/creatives and fighting for our rights that I created "Belladonna," which is inspired by the technology Nightshade. With that all being said: I am proud to reveal that this artwork is one of the first official Nightshaded artworks. The quote accompanying the painting Belladonna is "lumen inter tenebras," meaning light amid the darkness, a quote from Ben Y. Zhao. Nightshade will drop somewhere at the beginning of 2024 (for free) I recommend reading the entire article about it. I am linking it in my story. Belladonna is now available at WOW x WOW! Link in bio💓💕💕 #oilpaint #oilpainting #galleryart #notoai #notoaiart #illustration #illustrator #notoaigeneratedimages #beautifulbizarre #illustration #humanart #humanartist #artcommunity #lowbrow #creepycute

It’s been a big week for the Glaze project. We attended usenix security this week, where we presented the Glaze paper, and received both a Distinguished Paper Award as well as the 2023 USENIX Internet Defense Prize.

We wrapped up the week with a keynote at GNSI visualcomm, and an article about Glaze and artists in CNN.

'It gave us some way to fight back': New tools aim to protect art and images from AI's grasp | CNN Business For months, Eveline Fröhlich, a visual artist based in Stuttgart, Germany, has been feeling “helpless” as she watched the rise of new artificial intelligence tools that threaten to put human artists out of work.

We are thrilled to have our work again covered by the great Kashmir Hill at The New York Times. This is an upcoming paper to be presented next week at USENIX Security 2020.

https://www.nytimes.com/2020/08/03/technology/fawkes-tool-protects-photos-from-facial-recognition.html

You can run Fawkes on your own photos before uploading them to Facebook or other content sharing sites. It makes imperceptible tweaks to them, so that *if* a third party like Clearview.ai tried to use them to train a model against your face, they will fail. We designed the tool to be usable by anyone to protect their personal privacy against unauthorized surveillance and tracking. Link to download software binaries (or source code) in the comments.

Paper co-led by Shawn and Emily, with coauthors Huiying and Jiayun.

This Tool Could Protect Your Photos From Facial Recognition Researchers at the University of Chicago want you to be able to post selfies without worrying that the next Clearview AI will use them to identify you.

Students from the SAND Lab have two award (both best paper award honorable mentions) papers at the upcoming ACM CHI conference, the premier conference for human computer interaction research.

Jenna Cryan, Shiliang Tang, Xinyi Zhang led the paper “Detecting Gender Stereotypes: Lexicon vs. Supervised Learning Methods” where we looked at competing approaches to detect gender bias (lexicon vs deep learning), and find that deep learning methods, even when trained with minimal datasets, significantly outperform updated and augmented lexicon-based techniques
and
Yuxin Chen, Huiying Li and Shan-Yuan Teng led the paper “Wearable Microphone Jamming” where we showed how ultrasonic transducers embedded in a wearable bracelet can effectively jam nearby microphones (by inducing local responsible in the audible range via non-linearity), while relying on small wrist motions to reduce the impact of interference dead zones.

Congrats to all the authors!!!

Thrilled to announce that Huiying Li was just named a Facebook Fellow for 2020. She is one of only 36 fellows selected from over 1800 nominees! Way to go Huiying!!

Computer Science In just three years at the University of Chicago, PhD student Huiying Li has already studied security risks in machine learning, watermarks protecting deep neural networks, and technologies for “jamming” the microphones of nearby smart devices. That passion for privacy and security research set ...

Congratulations to Jenna Cryan, Yuxin Chen, Huiying Li, Shiliang Tang, Xinyi Zhang and Zhijing Li for their 2 papers at CHI 2020! 2/2 is a good year. See you in Honolulu!

What a Halloween we had this year! The lab theme was just everyone trying to look scary. Some succeeded beyond a doubt, while others just looked cute and goofy. In the end, Huiying Li won the best costume contest by a wide margin. Congrats Huiying, you scared us all! :D

Nothing like being thrown into the fire for your first public talk. Good luck to Huiying Li and Emily Willson as they travel to Stanford EE seminar to give their first technical talk on persistent and unforgeable DNN watermarks. Have fun!!

Super quick lab update!
It's almost the end of the summer! In the last week, we celebrated a birthday (Emily Willson turned 52 or 25, one of those), submitted 2 CHI papers (this afternoon, 1 by Jenna/Shiliang/Xinyi and 1 by Chen Yuxin/Huiying/Zhijing and Pedro Lopes). We saw the departure of the rest of our summer interns (Claudia and Alessia, and QinGe is leaving tomorrow).

Next week is the last week before classes start. New students Shinan and Zhuolin have been here for weeks, but they'll get "oriented" next week. Ben will be giving talks at Epfl Lausanne and ETH next week, and finishing up a paper submission (on the road) with Huiying and Emily to IEEE S&P!